MENTAL REPRESENTATIONS OF GOAL-DIRECTED PROCEDURES

正常认知的特点是能够灵活地学习新的程序和执行以前学习的程序。当这种能力受到干扰时，就像额叶失用症一样，后果是深远的。正常的程序性学习可能部分依赖于活动时间结构的分层组织表示。基于对事件表征的理论描述和最近关于目标定向任务知觉的研究，我假设这种表征在知觉和技能习得中是相同的。这里提出的研究使用程序性学习程序来检验这一假说。参与者将使用基于计算机的培训系统学习一项新颖的日常程序任务。通过改变计算机界面的布局，可以直接测试所建议的事件表示的作用：假设直接表示活动的时间组织的布局会带来更好的性能。通过改变呈现信息的媒介(静止图片或视频)，可以测试这些表示的特殊性。除了检验基本假设外，这两个拟议的实验都将检验一个次要的理论问题。在第一个实验中，对任务指令的外显记忆测试将表明外显记忆在多大程度上调节了对事件结构的理解。在第二个实验中，任务迁移测试将表明任务结构表征的内在化和灵活重复使用的程度。通过利用从最近的理论和实证工作中得出的范式，这些实验将阐明事件的结构化表征在正常认知中的作用。它们还提供了一个基础，以开发系统，以帮助患者处理受损的程序性知识，以及一个系统的基础，以协助技能指导。

Normal cognition is characterized by the ability to flexibly learn new procedures and to perform previously learned procedures. When this ability is disturbed, as in frontal apraxia, consequences are profound. Normal procedural learning may depend in part on hierarchically organized representations of the temporal structure of activity. Motivated by a theoretical account of event representations and recent work on the perception of goal-directed tasks, I hypothesize that such representations are shared across perception and skill acquisition. The studies proposed here test this hypothesis using a procedural learning program. Participants will learn a novel everyday procedural task, using a computer-based training system. Varying the layout of the computer interface allows a direct test of the role of the proposed event representations: Layouts that directly represent the temporal organization of the activity are hypothesized to lead to better performance. Varying the medium in which the information is presented (still pictures or video) allows a test of the specificity of these representations. In addition to testing the basic hypothesis, each of the two proposed experiments will examine a secondary theoretical issue. In the first experiment, a test of explicit memory for the task instructions will indicate the degree to which explicit memory mediates understanding of event structure. In the second experiment, a test of task transfer will indicate the degree to which representations of task structure are internalized and flexibly re-used. By taking advantage of a paradigm derived from recent theoretical and empirical work, these experiments will clarify the role of structured representations of events in normal cognition. They also provide a basis to develop systems to help patients cope with impaired procedural knowledge, as well as a basis for systems to assist in skill instruction.